Conventionally, communications between humans and machines have not been natural. Human-human communication typically involves spoken language combined with hand and facial gestures or expressions, and with the humans understanding the context of the communication. Human-machine communication is typically much more constrained, with devices like keyboards and mice for input, and symbolic or iconic images on a display for output, and with the machine understanding very little of the context. For example, although communication mechanisms (e.g., speech recognition systems) continue to develop, these systems do not automatically adapt to the activity of a user. As well, traditional systems do not consider contextual factors (e.g., user state, application state, environment conditions) to improve communications and interactivity between humans and machines.
Activity-centric concepts are generally directed to ways to make interaction with computers more natural (by providing some additional context for the communication). Traditionally, computer interaction centers around one of three pivots, 1) document-centric, 2) application-centric, and 3) device-centric. However, most conventional systems cannot operate upon more than one pivot simultaneously, and those that can do not provide much assistance managing the pivots. Hence, users are burdened with the tedious task of managing every little aspect of their tasks/activities.
A document-centric system refers to a system where a user first locates and opens a desired data file before being able to work with it. Similarly, conventional application-centric systems refer to first locating a desired application, then opening and/or creating a file or document using the desired application. Finally, a device-centric system refers to first choosing a device for a specific activity and then finding the desired application and/or document and subsequently working with the application and/or document with the chosen device.
Accordingly, since the traditional computer currently has little or no notion of activity built in to it, users are provided little direct support for translating the “real world” activity they are trying to use the computer to accomplish and the steps, resources and applications necessary on the computer to accomplish the “real world” activity. Thus, users traditionally have to assemble “activities” manually using the existing pieces (e.g., across documents, applications, and devices). As well, once users manually assemble these pieces into activities, they need to manage this list mentally, as there is little or no support for managing this on current systems.
All in all, the activity-centric concept is based upon the notion that users are leveraging a computer to complete some real world activity. Historically, a user has had to outline and prioritize the steps or actions necessary to complete a particular activity mentally before starting to work on that activity on the computer. Conventional systems do not provide for systems that enable the identification and decomposition of actions necessary to complete an activity. In other words, there is currently no integrated mechanism available that can dynamically understand what activity is taking place as well as what steps or actions are necessary to complete the activity.
Most often, the conventional computer system has used the desktop metaphor, where there was only one desktop. Moreover, these systems stored documents in a single filing cabinet. As the complexity of activities rises, and as the similarity of the activities diverges, this structure does not offer user-friendly access to necessary resources for a particular activity